Signal transduction cross talk mediated by Jun N-terminal kinase-interacting protein and insulin receptor substrate scaffold protein complexes

Abstract

Scaffold proteins have been established as important mediators of signal transduction specificity. The insulin receptor substrate (IRS) proteins represent a critical group of scaffold proteins that are required for signal transduction by the insulin receptor, including the activation of phosphatidylinositol 3 kinase. The c-Jun NH(2)-terminal kinase (JNK)-interacting proteins (JIPs) represent a different group of scaffold molecules that are implicated in the regulation of the JNK. These two signaling pathways are functionally linked because JNK can phosphorylate IRS1 on the negative regulatory site Ser-307. Here we demonstrate the physical association of these signaling pathways using a proteomic approach that identified insulin-regulated complexes of JIPs together with IRS scaffold proteins. Studies using mice with JIP scaffold protein defects confirm that the JIP1 and JIP2 proteins are required for normal glucose homeostasis. Together, these observations demonstrate that JIP proteins can influence insulin-stimulated signal transduction mediated by IRS proteins.

FIG. 1.

Proteomic analysis of JIP protein complexes. (A) The EE monoclonal antibody was used to prepare immunoprecipitates from Rin-5F cells (Control) and from Rin-5F cells expressing EE-tagged JIP1. The immunoprecipitates were examined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Silver-stained bands selectively present in JIP1 immunoprecipitates were excised and subjected to mass spectroscopy. Sixteen different proteins were identified. Eight of these correspond to subunits of the proteasome (listed below the gel). The identities of an additional eight proteins are annotated on the gel image. (B) Coimmunoprecipitation analysis of JIP1 complex formation. EE monoclonal antibody immunoprecipitates were prepared from Rin-5F cells (Control) and Rin-5F cells expressing EE-tagged JIP1. These immunoprecipitates were probed using antibodies to JIP proteins, JNK, ankyrin G, and ARP1. Immunoblot analysis of the cell lysate (100% of the amount used for the coimmunoprecipitation) is also presented.

FIG. 2.

Interaction of IRS and JIP proteins. (A) Immunoprecipitates of endogenous IRS1 and IRS2 were prepared from Rin-5F cells (Control) and Rin-5F cells expressing EE-tagged JIP1. These immunoprecipitates were probed using the EE monoclonal antibody and antibodies to IRS1 and IRS2. Immunoblot analysis of cell lysates (100% of the amount used for the coimmunoprecipitation) is also shown. (B) IRS proteins and T7 epitope-tagged JIP proteins were expressed in COS7 cells. The JIP proteins were immunoprecipitated using an antibody to the T7 epitope (IP). The amount of IRS and JIP proteins in the cell lysate and the immunoprecipitates was examined by immunoblot analysis using antibodies to IRS1, IRS2, and the T7 epitope tag on the JIP proteins. (C) Fragments of JIP1 with an NH₂-terminal T7 epitope tag were coexpressed with IRS2 in COS7 cells. The amount of IRS2 in the cell lysate and in JIP1 (T7) immunoprecipitates (IP) was examined by immunoblot analysis using an antibody to IRS2. The amount of JIP1 in the immunoprecipitates was examined by immunoblot analysis with a T7 monoclonal antibody. The heavy chain of the immunoglobulin used for immunoprecipitation is present in every lane and is indicated by an asterisk. (D) Hemagglutinin-tagged IRS1 proteins were coexpressed with T7-tagged JIP1 in COS7 cells. The effect of IRS1 mutations was examined, including the replacement of all 18 sites of Tyr phosphorylation with Phe (F18), deletion of the PH domain (ΔPH), deletion of the PTB domain (ΔPTB), and deletion of portions of the COOH-terminal region of IRS1 (PPY). JIP1 was immunoprecipitated using a T7 monoclonal antibody. The presence of IRS1 and JIP1 in the lysate and immunoprecipitate was examined by immunoblot analysis.

FIG. 3.

Interaction of endogenous JIP1 and IRS2. Rin-5F cells were serum-starved (16 h) followed by treatment with 100 nM insulin (A) or 10% serum (B). Lysates were prepared from the cells at the indicated times, and IRS2 was isolated by immunoprecipitation. The immunoprecipitates were examined by immunoblot analysis using antibodies to IRS2, phosphotyrosine, and JIP1. The amount of IRS2, JIP1, and JNK in the cell lysates was also examined by immunoblot analysis.

FIG. 4.

The diabetes-associated polymorphism (Ser-59Asn) in JIP1 suppresses JNK activation. (A) Lysates were prepared from Rin-5F cells expressing EE-tagged JIP1 or (S59N) JIP1. Lysates and JIP1 immunoprecipitates were probed by immunoblot analysis using antibodies to JIP1 and JNK. (B) Lysates were prepared from Rin-5F cells expressing EE-tagged JIP1 or (S59N) JIP1. Cell lysates and immunoprecipitates of endogenous IRS2 were probed by immunoblot analysis using antibodies to IRS2 and JIP1. (C) Wild-type Rin-5F cells and Rin-5F cells expressing EE-tagged JIP1 or (S59N) JIP1 were exposed to UV radiation (60 J/m²) and subsequently harvested at the indicated times. The amount of JIP1, JNK, and phospho-JNK was examined by immunoblot analysis (IB). JNK activity was measured in an immunocomplex kinase assay (KA) using [γ-³²P]ATP and c-Jun as substrates. An autoradiograph showing the phosphorylated c-Jun is presented, and the amount of phospho-c-Jun was quantitated by Phosphorimager analysis (mean ± standard deviation; n = 4). The JNK activity detected in cells expressing (S59N) JIP1 was significantly reduced compared with that of cells expressing wild-type JIP1 (P < 0.01).

FIG. 5.

Creation of mice with the germ line point mutation Ser-59Asn in the Jip1 gene. (A to B) A targeting vector was designed to replace Ser-59 with Asn and to introduce silent mutations that create an AseI site in exon II of the Jip1 gene by homologous recombination in ES cells. The floxed Neo^r cassette used for selection was deleted using Cre recombinase. (C) Genomic DNA isolated from wild-type and mutant ES cells was digested with AseI and examined by Southern blot analysis to confirm the correct targeting of the Jip1 gene. (D) Genomic DNA isolated from wild-type (Jip1^+/+), heterozygous (Jip1^+/S59N), and homozygous (Jip1^S59N/S59N) mice was examined by PCR analysis. (E) Brain lysates from Jip1^+/+ and Jip1^S59N/S59N mice were examined by immunoblot analysis using antibodies to JIP1, JIP2, JIP3, JIP4, and α-tubulin.

FIG. 6.

Metabolic effect of the Jip1^S59N mutation on the response to feeding a high-fat diet. Male Jip1^+/+ (WT) and Jip1^S59N/S59N (S59N) mice were fed a standard chow diet (control) or a high-fat (HF) diet for 0, 6, or 16 weeks. (A to B) The mice were fasted overnight, and the plasma concentrations of glucose (A) and insulin (B) were measured (mean ± standard deviation [SD]; n = 12). (C) The amount of plasma adiponectin, resistin, and leptin after 16 weeks on the diet was measured by multiplexed ELISA assays (mean ± SD; n = 12). (D) The activity of JNK in epididymal fat pads of the mice after 16 weeks on the control or HF diet was measured by using the immunocomplex kinase assay (KA) using [γ-³²P]ATP and c-Jun as the substrates. The amount of JNK and α-tubulin in the cell lysate was examined by immunoblot analysis.

FIG. 7.

Jip1/2 compound mutant mice exhibit severe growth retardation. (A and B) Littermate Jip1^−/− Jip2^−/− mice (left) and control (Jip1^−/+ Jip2^−/+) mice (right) at postnatal day 2 and day 10 are illustrated. (C) The body masses of littermate Jip1^−/− Jip2^−/− mice and control (Jip1^−/+ Jip2^−/+) mice at postnatal day 10 are presented (mean ± standard deviation [SD]; n = 6). The masses of the Jip1^−/− Jip2^−/− mice were significantly less than those of the control mice. *, P < 0.001. (D and E) Blood collected from wild-type and Jip1^−/− Jip2^−/− mice was used to measure the amount of glucose (D) and insulin (E). The data presented are the mean ± SD (n = 10). Significant differences between the Jip1^−/− Jip2^−/− mice and control mice are indicated. *, P < 0.01.

FIG. 8.

Defects in the phosphorylation of IRS1 on Ser-307 in Jip1/2 compound mutant mice. Wild-type mice, Jip1^−/− mice, and Jip1^−/− Jip2^−/− mice (postnatal day 1) were treated without and with 1.5 U/kg insulin by intraperitoneal injection. Extracts prepared from brown fat were examined by immunoblot analysis using antibodies to IRS1 and α-tubulin. IRS1 immunoprecipitates prepared from the cell lysate were probed with antibodies to IRS1 and phospho-Ser-307 IRS1. Densitometric analysis of immunoblots was employed to quantitate the effect of insulin to increase the phosphorylation of IRS1 on Ser-307.

FIG. 9.

Pancreatic morphology of Jip1/2 compound mutant mice. (A to C) Sections of the pancreases from 7-day-old wild-type and Jip1^−/− Jip2^−/− mice were stained with hematoxylin and eosin (A) or with antibodies to insulin (B) or to insulin and glucagon (C). (D) The morphology of pancreatic β cells was examined by using electron microscopy.